Contributions of nuclear cardiology to diagnosis and prognosis of patients with coronary artery disease.

Abstract

HomeCirculationVol. 101, No. 12Contributions of Nuclear Cardiology to Diagnosis and Prognosis of Patients With Coronary Artery Disease Free AccessOtherPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyRedditDiggEmail Jump toFree AccessOtherPDF/EPUBContributions of Nuclear Cardiology to Diagnosis and Prognosis of Patients With Coronary Artery Disease George A. Beller and Barry L. Zaret George A. BellerGeorge A. Beller From the Cardiovascular Division, University of Virginia Health System, Charlottesville, Va (G.A.B.), and the Cardiology Division, Yale University, New Haven, Conn (B.L.Z.). Search for more papers by this author and Barry L. ZaretBarry L. Zaret From the Cardiovascular Division, University of Virginia Health System, Charlottesville, Va (G.A.B.), and the Cardiology Division, Yale University, New Haven, Conn (B.L.Z.). Search for more papers by this author Originally published28 Mar 2000https://doi.org/10.1161/01.CIR.101.12.1465Circulation. 2000;101:1465–1478In the past decade, significant advances have been made in the ability to image the heart with radionuclide tracers under stress and resting conditions in patients with suspected or known coronary artery disease (CAD) for the detection of ischemia, determination of prognosis, assessment of myocardial viability, preoperative risk assessment for patients undergoing noncardiac surgery, and evaluation of the efficacy of revascularization in patients undergoing coronary artery bypass surgery or an interventional procedure.1For many years, planar imaging and SPECT with 201Tl constituted the only scintigraphic techniques available for detecting CAD and assessing prognosis in patients undergoing stress perfusion imaging. The major limitation of 201Tl scintigraphy is the high false-positive rate observed in many laboratories, which is attributed predominantly to image attenuation artifacts and variants of normal that are interpreted as defects consequent to a significant coronary artery stenosis. Although quantification of 201Tl images improves specificity, the false-positive rate remains problematic, particularly in women and in obese patients. Breast attenuation artifacts in women are sometimes difficult to distinguish from perfusion abnormalities secondary to inducible ischemia or myocardial scar.In recent years, new 99mTc-labeled perfusion agents have been introduced into clinical practice to enhance the specificity of SPECT and to provide additional information regarding regional and global left ventricular systolic function via ECG gating of images. It was immediately apparent that the quality of images obtained with these new 99mTc-labeled radionuclides was superior to that of images obtained with 201Tl because of the more favorable physical characteristics of 99mTc imaging with a gamma camera. With 99mTc, doses of ≈10 to 20 times higher than those that are feasible with 201Tl can be administered, yielding images with higher count density. 99mTc demonstrates less scatter and attenuation than 201Tl, which is associated with fewer image artifacts in patients with no underlying CAD. Perhaps most importantly, 99mTc-sestamibi or 99mTc-tetrofosmin imaging allows easy gated acquisition, permitting the simultaneous evaluation of regional systolic thickening, global left ventricular function, and myocardial perfusion.2 Experimental studies have shown that the uptake of these new 99mTc tracers is proportional to regional blood flow but plateaus earlier than 201Tl at hyperemic flows because extraction is reduced. It should be pointed out that the first-pass myocardial extraction fraction of 99mTc-tetrofosmin is lower than 99mTc-sestamibi by ≈10% to 15%, which may contribute to its lower sensitivity for detecting mild to moderate stenoses with vasodilator stress. A new 99mTc-labeled perfusion agent, 99mTc-N-NOET, presently under investigation shows a higher first-pass myocardial extraction than 99mTc-sestamibi or 99mTc-tetrofosmin and redistributes over time similar to 201Tl.3Diagnosis of CADExercise Perfusion ImagingPerhaps 1 of the most significant advances in myocardial perfusion imaging in the past decade is the development of quantitative SPECT perfusion imaging. With planar imaging incorporating visual assessment of myocardial scintigrams, sensitivity and specificity for detection of CAD averaged 82% and 88%, respectively, in ≈4000 patients combined from multiple published series.4 The sensitivity and specificity of quantitative planar 201Tl scintigraphy were significantly higher at 91% and 89%, respectively, in 682 patients from studies published in the literature.5 With tomographic SPECT technology, sensitivity averaged 92% (range, 82% to 98%) with a specificity of 68% in 1447 patients combined from 6 studies in the literature. The specificity for CAD detection was only 68% (range, 44% to 91%), which is most likely attributed to a referral bias in which patients with abnormal scans are more likely to be referred for coronary angiography than patients with normal scans. The normalcy rate was 84%. The normalcy rate, used as a surrogate for specificity, is defined as the rate of normal perfusion scans in patients with <5% likelihood of CAD on the basis of clinical and ECG stress test data. It is of interest that the sensitivity of SPECT remains high (85%) for CAD detection in patients without prior myocardial infarction. As expected, the sensitivity for detection of single-vessel disease with SPECT 201Tl imaging averaged 83% compared with 93% for patients with angiographic 2-vessel disease and 95% for patients with 3-vessel disease.4Variables that diminish sensitivity of SPECT 201Tl imaging for CAD detection are single-vessel disease, left circumflex coronary artery stenosis, branch vessel or distal stenosis, mild degree of stenosis (50% to 70% luminal narrowing), inadequate heart rate response with cessation of exercise caused by noncardiac symptoms, and antianginal therapy with nitrates or calcium blockers. As expected, sensitivity is enhanced in patients with prior myocardial infarction, more extensive CAD, high-grade coronary stenosis, proximal location of stenosis, and presence of regional wall motion abnormalities. Quantitative scan analysis yields higher sensitivity and specificity values than visual assessment alone of stress and redistribution 201Tl scintigrams. In 1 study,6 stenosis severity was the most important determinant of an abnormal scan, followed by impaired treadmill exercise duration caused by cardiac symptomatology.As indicated above, the overall specificity of 201Tl scintigraphy is suboptimal, which is due predominantly to a failure to recognize image attenuation artifacts. Although quantification of 201Tl images improves specificity, the false-positive rate still remains unacceptable. The false-positive defects are commonly observed in the inferoapical region toward the basilar segment of the left ventricle. In women, attenuation artifacts are caused by overlying breast tissue and are localized in the anterior wall and septum. A high diaphragm can cause attenuation artifacts in the inferior wall. Gated 99mTc perfusion imaging permits the assessment of systolic thickening of end diastole to end systole on SPECT tomograms. Recognition of normal systolic thickening in an area of hypoperfusion would direct the interpreter to the conclusion that a defect is an attenuation artifact rather than a myocardial scar. The latter would most likely be associated with reduced systolic thickening as assessed on gated tomograms. In a prospective study by Taillefer et al,7 the diagnostic accuracy of 201Tl SPECT and 99mTc SPECT perfusion imaging for detection of CAD in women was evaluated. Women in the study underwent both SPECT techniques, and many underwent coronary angiography. The overall sensitivities for detecting significant CAD were similar for 201Tl SPECT and 99mTc SPECT. The specificity, however, was only 67% for 201Tl SPECT when the definition of a significant stenosis was ≥70% reduction in luminal diameter (Figure 1). In the same women, 99mTc-sestamibi SPECT perfusion imaging alone increased specificity to 84%. When gated images were then also analyzed, specificity for CAD detection further improved to 92%. Thus, gated 99mTc-sestamibi imaging reduced the false-positive rate in these women from 33% with 201Tl to 8% with gated SPECT using 99mTc-sestamibi.A study performed in a heterogeneous population of patients also demonstrated the value of gating of 99mTc-sestamibi SPECT images.8 This study showed that the addition of gating to standard perfusion 99mTc-sestamibi SPECT reduced the number of borderline interpretations from 89 to 29 in the total group of 285 patients. In the 137 patients with a pretest likelihood of CAD of ≤10%, the addition of gated images added significantly to the percentage of interpretations that were designated as “normal” (74% to 93%) because of a reduction in “borderline-normal” and “borderline-abnormal” readings. Thus, the addition of ECG-gated 99mTc SPECT images to the reading of stress and rest perfusion images alone resulted in a significant shift in the final scan interpretations to a more normal designation in patients with a low pretest likelihood of CAD and to more abnormal readings in patients with known CAD.Finally, ECG-gated SPECT yields important information about global left ventricular function that could previously be obtained only with a second test, such as radionuclide angiography, echocardiography, or contrast ventriculography.910 The ability to accurately measure left ventricular ejection fraction with 99mTc-sestamibi or 99mTc-tetrofosmin adds supplementary value to the procedure, particularly with respect to prognostication and assessment of viability.A pooled analysis of studies using exercise 99mTc-sestamibi SPECT imaging yielded a 90% sensitivity for 99mTc-sestamibi and 83% sensitivity for 201Tl for the detection of CAD.11 Specificity for SPECT 99mTc-sestamibi was 93% compared with 80% for SPECT 201Tl, whereas the normalcy rate was 100% for 99mTc-sestamibi and 77% for 201Tl. In a phase III multicenter SPECT trial, detection of single-vessel disease was 90% for 99mTc-sestamibi imaging, which was ≈20% higher than the detection rate of single-vessel disease with the planar imaging approach in this phase III trial.11 Sensitivity for detecting 3-vessel disease was 98% for 99mTc-sestamibi SPECT imaging.An alternative approach to performing rest and stress 99mTc-sestamibi imaging for detection of CAD and distinguishing reversible from nonreversible defects is a dual-isotope rest 201Tl/stress 99mTc-sestamibi SPECT imaging protocol. In this protocol, 3.5 mCi of 201Tl is injected at rest, with images acquired 10 minutes later. Exercise testing is performed immediately after acquisition of these rest images with 25 to 30 mCi of 99mTc-sestamibi injected at peak stress. Berman and coworkers12 reported a sensitivity and specificity of 91% and 75%, respectively, for CAD detection with this technique. The normalcy rate in patients with a low pretest likelihood of CAD was 95%. Defect reversibility, in a comparison of the resting 201Tl and 99mTc-sestamibi scintigrams, was comparable to the prevalence of defect reversibility on the standard rest/stress 99mTc-sestamibi SPECT technique.Pharmacological Stress Perfusion ImagingPharmacological stress imaging is an important alternative noninvasive approach for detecting CAD and assessing prognosis in patients who are unable to adequately exercise. Intravenous infusion of dipyridamole, adenosine, or dobutamine is an acceptable alternative to an exercise stress for determining the presence and extent of CAD using SPECT perfusion imaging with 201Tl or 99mTc-sestamibi. Both dipy-ridamole and adenosine induce a 3- to 5-fold increase in myocardial blood flow with the standard doses. Although adenosine causes a greater increase in blood flow and more patients achieve a maximal flow increase than with dipy-ridamole, both adenosine and dipyridamole increase blood flow in normal myocardium in excess of that found with maximal exercise. Both have comparable detection rates for identifying functionally important coronary stenoses. The sensitivity for CAD detection for dipyridamole, adenosine, and dobutamine ranges from 89% to 91%.13 Specificity is also comparable among these 3 stressors, although the specificity for CAD detection might be slightly higher for adenosine compared with dipyridamole. Recent data14 suggest that in women adenosine perfusion imaging is significantly more sensitive than exercise imaging for detection of single-vessel disease. In men, the sensitivity and specificity for CAD detection were comparable for exercise and adenosine SPECT perfusion imaging with comparable specificities.Patients with left bundle-branch block (LBBB) and angiographically normal coronary arteries often have abnormal septal defects on exercise SPECT perfusion imaging. The false-positive rate of septal defects in patients with LBBB for CAD detection is significantly lower with dipyridamole or adenosine perfusion imaging. Therefore, in such patients, vasodilator stress imaging is preferable to exercise imaging to determine whether CAD causes the LBBB pattern.Side effects are somewhat greater with adenosine vasodilator stress imaging than with dipyridamole imaging. Of 9256 patients undergoing adenosine perfusion imaging, 82% had adverse side effects, the most common of which were flushing (37%), chest pain (35%), shortness of breath or dyspnea (35%), headache (14%), ECG ischemic changes (9%), and 8-V conduction block (8%).15 New adenosine A2A-receptor agonists are under investigation for clinical use instead of dipyridamole or adenosine for vasodilator stress imaging. These A2A-receptor agonists selectively dilate the coronary arteries without associated systemic hypotension or without the adenosine A1-receptor agonist effects, such as AV block and nonischemic chest pain.16Intravenous dobutamine infusion is an alternative to vasodilator stress for myocardial perfusion imaging and is predominantly indicated in patients with pulmonary disease and bronchospasm. Dobutamine infusion produces flow heterogeneity in the presence of a significant coronary artery stenosis because it increases myocardial oxygen demand by increasing heart rate, blood pressure, and contractility. O’Keefe et al13 summarized the published studies in which dobutamine was used with perfusion scintigraphy. Sensitivity and specificity for CAD detection were 91% and 86%, respectively, in the 158 patients included in these 3 studies. Hays et al17 found that the sensitivity for single-, double-, and triple-vessel CAD was 84%, 82%, and 100%, respectively, in a group of patients undergoing SPECT dobutamine 201Tl imaging. Overall sensitivity for CAD detection was 86%, with a 90% specificity.Exercise or pharmacological stress myocardial perfusion imaging provides significant supplementary diagnostic information to stress ECG variables for detecting CAD among patients presenting with chest pain. The reason is that both the sensitivity and specificity of exercise ECG stress testing are suboptimal for CAD detection. Gianrossi et al18 performed a meta-analysis of 147 published studies in the literature in which the exercise ST-segment response was compared with coronary angiographic findings. The mean sensitivity for detection of CAD in these studies was 68%, with a specificity of 77%. The extent of CAD certainly affects the sensitivity of the exercise ST-segment response. Sensitivity for detection of single-vessel disease with ECG stress testing alone ranges from 50% to 55%. Sensitivity is also significantly reduced in patients who are unable to achieve ≥85% of their maximum predicted heart rate for their age. In the presence of baseline ECG abnormalities, exercise-induced ST-segment depression may be nonspecific for ischemia, which considerably lowers the specificity of the test.Prognostic Value of Stress Perfusion ImagingThe prognostic value of exercise and pharmacological stress perfusion imaging has been established in thousands of patients evaluated in multiple clinical studies.19 The major goal of noninvasive risk stratification with stress perfusion imaging in either patients presenting with undiagnosed chest pain or patients with known CAD is the identification of subsets at high risk of cardiac death or nonfatal infarction so that prompt referral to invasive strategies can be undertaken. Conversely, patients at low risk of future cardiac events on the basis of scintigraphic findings can be spared unnecessary referral for invasive evaluation.The major prognostic variables on stress perfusion images predictive of future cardiac events are a large defect size (>20% of the left ventricle), defects in >1 coronary vascular supply region suggestive of multivessel CAD, defect reversibility reflective of inducible ischemia in multiple myocardial scan segments, a large number of nonreversible defects even in the supply region of a single coronary artery, transient or persistent left ventricular cavity dilation from stress to rest images, increased lung 201Tl uptake on 201Tl scintigraphy, and a resting left ventricular ejection fraction measured on 99mTc gated SPECT imaging of <40%.Perhaps 1 of the most valuable features of exercise or pharmacological stress perfusion imaging with 201Tl or a 99mTc-labeled agent is its excellent negative predictive value for predicting low mortality and myocardial infarction rates in patients with totally normal scans. Patients with normal perfusion studies at peak stress have a <1%/y combined mortality and nonfatal infarction rate and are thus often spared further invasive evaluation for assessment of their symptoms.20One of the first published reports of the prognostic value of exercise 201Tl imaging by Brown et al21 revealed that the number of reversible 201Tl defects was the best predictor of death or nonfatal infarction in 100 patients without prior myocardial infarction. In that study, neither the number of stenotic arteries on angiography nor the presence of inducible ST-segment depression on exercise electrocardiography provided significant supplementary prognostic information to the scintigraphic data. Similarly, Ladenheim et al22 identified the number of myocardial segments with reversible 201Tl defects and exercise heart rate as the only independent predictors of future cardiac events in 1689 CAD patients without prior infarction. Other studies with either planar or SPECT 201Tl scintigraphy confirmed these earlier studies and showed that when assessment of variables for myocardial perfusion imaging is added to information solely obtained from the clinical history, physical examination, and exercise treadmill results, incremental prognostic information is obtained.2324 Perhaps the largest prognostic study with SPECT 201Tl imaging is by Machecourt et al,25 who followed up 1926 patients for 33 months after exercise 201Tl SPECT imaging. The cardiac mortality rate was 0.42%/y in patients with normal scans and 2.1%/y in patients with abnormal scans. When ≥3 scan territories were involved, the relative risk of cardiac death increased to 24.The prognostic value of stress 99mTc-sestamibi perfusion imaging is comparable to that reported with 201Tl imaging. Iskander and Iskandrian26 analyzed 14 prognostic studies comprising more than 12 000 patients with respect to the prognostic value of the perfusion imaging data. In these patients, normal stress SPECT 99mTc-sestamibi images were associated with an average annual hard event rate of 0.6%. Patients with abnormal images had a 12-fold-higher event rate at 7.4% annually (Figure 2). Berman et al27 demonstrated the incremental prognostic value of exercise 99mTc-sestamibi imaging in 1702 patients who were followed up after testing for 20±5 months. In this study, patients were divided into low, intermediate, and high pretest likelihood of CAD on the basis of clinical and exercise ECG stress test data. 99mTc-sestamibi scan results further separated patients into lower and higher subgroups. A very important finding in this article was that the cardiac event rate for patients with normal scans was low for all levels of pretest likelihood of CAD after acquisition of exercise ECG stress test results. For example, of the 90 patients who were classified as having a >85% likelihood of CAD after the exercise treadmill test and who had a normal scan, none had a cardiac event during follow-up. The rates of referral for catheterization after 99mTc-sestamibi SPECT imaging was also very low in patients with normal perfusion studies, even if they had a high likelihood of CAD when only the clinical and exercise stress test results were analyzed.In another publication from this group, Hachamovitch et al28 showed that the catheterization rate was only 1% in 834 patients who had an intermediate Duke treadmill score after exercise testing but a low-risk 99mTc-sestamibi stress perfusion scan. Interestingly, the combined death and infarction rate in 834 patients with an intermediate Duke treadmill score and a normal scan was 0.4%/y. In contrast, those with an intermediate Duke treadmill score and a high-risk SPECT scan had an 8.9% annual combined cardiac death and nonfatal infarction rate (Figure 3). Exercise 99mTc-sestamibi perfusion imaging variables provide even greater incremental prognostic value for women than men.2930 In the study by Hachamovitch et al,29 receiver-operating characteristic analysis demonstrated superior discrimination for the nuclear scan results in identifying high-risk women than in men (area under curve, 0.84 in women versus 0.71 in men). In the study by Marwick et al,30 the number of abnormal territories in the distribution of the 3 major coronary arteries remained the strongest correlate of mortality after adjustment for exercise variables in 3402 women analyzed. Boyne et al31 determined the ability of exercise 99mTc-sestamibi SPECT imaging to predict adverse events in a population with a considerable number of women. In patients with normal scans, the combined cardiac death and nonfatal infarction rate was 0.8%/y compared with a 5.4%/y event rate in patients with abnormal scans.Good exercise tolerance on treadmill testing is associated with a favorable prognosis. However, even in patients who achieve stage IV or greater on the Bruce protocol, exercise myocardial perfusion imaging provides significant additional prognostic value.32 In such patients, Chatziioannou et al32 reported that with the use of Cox proportional-hazards regression analysis, myocardial perfusion imaging was an excellent predictor of cardiac events (global χ2=13.2; P<0.001; relative risk=8), but exercise ECG variables or the addition of the Duke treadmill score risk categories had no predictive power (Figure 4).The extent of hypoperfusion on poststress 99mTc-sestamibi images can be factored into a decision-making process relative to selecting medical therapy or revascularization. Patients with mild reversible perfusion defects judged to be not high risk can most often be treated medically, whereas patients with high-risk SPECT reversibility findings are candidates for further invasive strategies. O’Keefe et al33 undertook a follow-up study in patients with mild to moderate SPECT reversibility, only 9% of whom were initially referred for coronary angiography. With medical therapy, the unadjusted actuarial 3-year event rate for cardiac death or nonfatal infarction was only 2% in the medically managed patients with non–high-risk SPECT reversibility. Only 4% of patients crossed over to subsequent coronary revascularization during follow-up. This study suggested that the results of SPECT imaging assisted in deciding which patients could do well with initial medical therapy as reflected by the low mortality and infarction rates at 3 years of follow-up.Hachamovitch et al34 also found a low annual cardiac death rate (0.8%/y) in patients with mildly abnormal stress perfusion scans who received medical therapy compared with an annual cardiac death rate of 0.9% in patients with mildly abnormal scans who underwent revascularization. This was an observational retrospective analysis of patients who underwent either medical or revascularization therapy after stress 99mTc-sestamibi testing. In that study, patients with moderately abnormal or severely abnormal scans appeared to benefit from revascularization early after nuclear testing (Figure 5).Pharmacological stress imaging can also be used for prognostication in patients with chronic CAD and in asymptomatic patients with aortic or peripheral vascular disease who are scheduled for major vascular surgery. High-risk imaging variables are comparable to those described previously for exercise perfusion imaging. Table 1 shows the prognostic value of adenosine-stress SPECT imaging in 1159 consecutive patients.35 Patients with a low summed stress score had a cardiac death rate of only 0.9%/y and a combined death and nonfatal infarction rate of 1.6%/y. As the summed stress score became more severe, the cardiac event rate substantially increases, rising to a mortality rate of 7.4% with a severely abnormal scan.Thus, taken together, these data reported from the literature demonstrate that patients with normal myocardial perfusion scans have an excellent prognosis even if angiographic CAD is documented, whereas patients with abnormal scans have an increased rate of cardiac death and nonfatal infarction during follow-up. The greater the extent of stress-induced hypoperfusion and reversibility, the greater is the probability of an event. Similarly, the event rate is higher for any extent of hypoperfusion if transient ischemic left ventricular dilation or increased lung 201Tl uptake is observed. A strategy incorporating stress myocardial perfusion imaging as the initial test for detecting CAD and assessing prognosis in patients with chest pain and suspected CAD is also cost-effective.A large observational study36 comprising 11 372 consecutive stable angina patients referred for stress myocardial perfusion SPECT imaging or direct catheterization revealed that costs were higher for the initial invasive strategy in clinical subsets with low, intermediate, or high pretest likelihood of disease (see Figure 6). Diagnostic and follow-up costs of care were 30% to 41% higher for patients undergoing direct cardiac catheterization without any reduction in mortality or infarction. The diagnostic costs were $1320, $1275, and $1229 greater for low-, intermediate-, and high-risk patients undergoing initial cardiac catheterization compared with those having stress perfusion imaging as the initial test for CAD detection. The cardiac death rate and nonfatal infarction rate in the 5826 patients undergoing initial stress perfusion imaging for assessment of stable angina were both 2.8% compared with 3.3% and 3.0%, respectively, for the 5423 patients who were referred directly for cardiac catheterization as the initial diagnostic strategy. Thus, stress myocardial perfusion imaging may serve as a gatekeeper for referral for cardiac catheterization. This strategy, in which referral for cardiac catheterization is “ischemia driven,” could reduce healthcare costs because coronary angiography could be avoided in patients with low-risk radionuclide perfusion studies regardless of clinical characteristics, treadmill exercise test results, and even coronary angiographic findings. The latter is possible because of the low risk for subsequent cardiac death and infarction in patients with demonstrated angiographic coronary disease and normal perfusion scans.37 Figure 7 is a proposed decision-making algorithm for stable patients with an intermediate or high pretest likelihood of CAD who are referred for stress SPECT imaging, which is derived from the prognostic data summarized above.Risk Stratification After Acute Myocardial InfarctionRadionuclide techniques can be clinically useful in the evaluation of patients who have had an acute myocardial infarction with respect to determining infarct size, assessing the degree of myocardial salvage after reperfusion, determining myocardial viability in infarct zones of resting asynergy, and detecting inducible myocardial ischemia within or remote from the infarct zone with exercise or pharmacological stress.3839404142High-risk myocardial perfusion imaging variables in patients undergoing stress scintigraphy before discharge after acute myocardial infarction are (1) reversible defects within the infarct zone, (2) a multivessel disease scan pattern, (3) a large nonreversible defect corresponding to a large infarct size or cumulative zone of nonviability (new and old infarction), (4) transient left ventricular cavity dilation from stress to rest imaging, (5) increased lung/201Tl uptake when that tracer is used, and (6) resting ejection fraction <40% on gated SPECT. A pooled analysis performed by Shaw et al43 combining studies relevant to exercise stress myocardial perfusion imaging after myocardial infarction showed the mortality rate to be 7.1% in patients with a stress-induced reversible defect on stress imaging compared with 1.6% in those without a reversible defect. Similarly, patients with multiple defects in >1 coronary supply region had a 16.7% combined death or myocardial infarction rate compared with a 2% event rate in patients without a multivessel disease scan pattern. In the era before thrombolytic therapy, Gibson et al44 used submaximal exercise 201Tl scintigraphy before hospital discharge and demonstrated that ≈50% of patients with an uncomplicated myocardial infarction who had a high-risk scan experienced a subsequent cardiac event (cardiac death, nonfatal infarction, or rehospitalization fo